IMPACT OF SIGNATURE LEGIBILITY AND SIGNATURE TYPE IN

OFF-LINE SIGNATURE VERIFICATION

F Alonso-Fernandeza, M C. Fairhurstb, J Fierrez' and J Ortega-Garciaa.

aBiometric Recognition Group - ATVS, Escuela Politecnica Superior - Universidad Autonoma de Madrid
Avda. Francisco Tomas y Valiente, 11 - Campus de Cantoblanco - 28049 Madrid, Spain
{ fernando.alonso, julian.fierrez, javier.ortega}@uam.es

bDepartment of Electronics, University of Kent, Canterbury, Kent CT2 7NT, UK

{ M.C.Fairhurst} @kent.ac.uk

readable written name (e.g. American signatures). In other

The performance of two popular approaches for off-line sig- cases, as frequently happens in European countries, signatures may consist of only an elaborated flourish. In contrast
nature .f i isigto occidental signatures, oriental signatures consist of indei
a
d
nature
t
nature type is studied. We investigate experimentally if the' pendent symbols.
Examples of both oriental aland occidental
hes Frth inte
Signtat
knowledge of letters, syllables or name instances can help in signatuscabeou
the process of imitating a signature. Experimental results are
..
given on a sub-corpus of the MCYT signature database for Verification Competition [5].
Signature verification systems have been shown to be senrandom and skilled forgeries. We use for our experiments
sitive
to some extent to signature complexity [6]. Easy to
two machine experts, one based on global image analysis and
forge signatures result in increased False Acceptance Rate.
statistical distance measures, and the second based on local
Signature variability also has an impact in the verification
image analysis and Hidden Markov Models. Verification rerates
attainable [7]. It can be hypothesized that these two
in
sults are reported terms of Equal Error Rate (EER), False
1
factors,
complexity and variability, are related in some way
Rate
and
False
Rate
(FAR)
(FRR).
Acceptance
Rejection
with signature legibility and signature type. Moreover, some
studies have been concerned with the ability ofhumans in rec1. INTRODUCTION
ognizing handwritten script [8, 9]. Knowledge about letters,
syllables or name instances may help in the process of imitatThe handwritten signature is one of the most widely used ining a signature, which is not the case for an incomprehensible
dividual authentication methods due to its acceptance in govset of strokes that, in principle, are not related to any linguistic
emnment, legal and commercial transactions as a method of knowledge.
identity verification [ 1, 2]. As a result, a number of algoThe main goal ofthis work is to evaluate the impact of sigrithms have been proposed for automatic signature verificanature legibility and signature type on the recognition rates of
tion [3]. This work is focused on off-line verification, a pat- two popular approaches to off-line signature verification. In
tern classification problem with a long history, involving the this paper, signature legibility and type are assessed by a hudiscrimination of signatures written on a piece of paper [4].
man expert. Some examples are shown in Figs. 1 and 2. This
It is worth noting that even professional forensic document
process is not unreasonable in relation to off-line signature
examiners perform a correct classification rate of only about verification environments, where signature acquisition is typ70%, confirming that this a challenging research area.
ically performed by a human operator using a scanner or a
In this paper, we focus on occidental signatures, which
camera [4].
typically consist of connected text (i.e. name) and/or some
Two machine experts with different approaches for feaform of flourish. Sometimes, signatures only consist of a ture extraction are used in the work reported here, as deABSTRACT

vyperificatudion Wer oestignature lxpegibimealty

1This work has been carried out while F. A.-F. was guest scientist at
the University of Kent. This work has been supported by Spanish MCYT
TEC2006-13141-C03-03 and by European Commission IST-2002-507634
Biosecure NoE projects. Author F. A.-F. thanks Consejeria de Educacion
de la Comunidad de Madrid and Fondo Social Europeo for supporting his
PhD studies. Author J. F. is supported by a Marie Curie Fellowship from the

European Commission.

1-4244-1 549-7/07/$25.OO

2007 IEEE

scribed in Section 2. The first is based on global image anal-

ysis and a minimum distance classifier as proposed in [10],

and further developed in [11]. The second is based on local image analysis and left-to-right Hidden Markov Models

as used in [12] but with a local parameterization derived from

[10], and also detailed in [11]. The rest of this paper is orga-

2007 Biometrics Symposium

NAME NO LEGIBLE OR NO NAME

MEDIUM LEGIBILITY

NAME CLEARLY LEGIBLE

Fig. 1. Signature examples with different degrees of name legibility (from top to bottom).
SIMPLE FLOURISH

COMPLEX FLOURISH

NAME + SIMPLE
FLOURISH

NAME + COMPLEX
FLOURISH

Fig. 2. Signature examples of the four types encountered in the MCYT corpus (from left to right).

nized as follows. The experimental framework used, including the database, protocol and results, is described in Section 3. Some conclusions are finally drawn in Section 4.
2. MACHINE EXPERTS
In this section, the two machine experts used in this paper
are described. They exploit information at two different levels: the first approach analyze the image in a holistic manner,
wheres the second approach is based on features extracted locally. Additional details can be found in [ 1].

2.1. Based on global information

Input signature images are first preprocessed according to the
following consecutive steps: binarization by global thresholding of the histogram [13], morphological closing operation on
the binarized image [14], segmentation ofthe signature outer
traces, and normalization of the image size to a fixed width of
512 pixels while maintaining the aspect ratio (see Fig. 3 for an
example). Normalization ofthe image size is usedto make the
proportions of different realizations of an individual sample to

1-4244-1 549-7/07/$25.OO

2007 IEEE

be the same, whereas segmentation of the outer traces is carried out because a signature boundary typically corresponds
to a flourish, which has high intra-user variability. For this
purpose, left and right height-wide blocks having all columns
with signature pixel count lower than threshold Tp and top
and bottom width-wide blocks having all rows with signature
pixel count lower than Tp are discarded.
A feature extraction stage is then performed, in which
slant directions of the signature strokes and those of the envelopes of the dilated signature images are extracted using
mathematical morphology operators [14], see Fig. 4. These
descriptors are used as features for recognition as proposed in
[10]. For slant direction extraction, the preprocessed signature image is eroded with 32 structuring elements, thus generating 32 eroded images. A slant direction feature sub-vector
of 32 components is then generated, where each component is
computed as the signature pixel count in each eroded image.
For envelope direction extraction, the preprocessed signature
image is successively dilated 5 times with each one of 6 linear structuring elements, thus generating 5 x 6 dilated images.
An envelope direction feature sub-vector of 5 x 6 components
is then generated, where each component is computed as the

2007 Biometrics Symposium

BINARIZA11N

CLOSINGY

Vertical

pixorizntalpixensirtit

Fig. 3. Preprocessing stage performed in the global expert.

-level________

_________
signature pixel count in the difference image between suc-Lgblt
ubro sr
cessive dilations. The preprocessed signature is finally paLgblt ee
1usr(2%
rameterized as a vector o with 62 components by concatenatNon-legible
Medium
19ues(53%
ing the slant and envelope feature sub-vectors. Each client
Legible
38ues(06%
(enrolee) of the system is represented by a statistical model
A =4/t, a-) which is estimated by using an enrolment set of
_________
Type________
yeNubrofsr
K parameterized signatures {Oi, OK}. The parameters /
Simple flourish
5 sr 66%
and a- denote mean and standard deviation vectors of the K
Complex flourish
vectors {Oi, OK}. In the similarity computation stage, the
13ues(73%
Name + simple flourish
5ues(66%
similarity score between a claimed model A =(iit, a-) and a
+ complex flourish
Name
2ues(93%
parameterized test signature o is computed as the inverse of
the Mahalanobis distance [15].
Table 1. Distribution of users on the MCYT database based
on name legibility and signature type.
2.2. Based on local information
..

..

In the preprocessing stage, images are first binarized and segmented as described in Section 2.1. Next, afeature extraction
step is performed, in which slant directions and envelopes are
locally analyzed using the approach described in Section 2. 1,
but applied to blocks. Preprocessed images are divided into
height-wide blocks of 64 pixels width with an overlapping
between adjacent blocks of 75%. The rightmost block is discarded. A signature is then parameterized as a matrix 0 whose
columns are 62-tuples, each one corresponding to a block,
Each client of the system is represented by a Hidden Markov
Model A (HMM) [ 16, 17], which is estimated by using an enrolment set of K parameterized signatures {O0 1, ...,~OK}. A
left-to-right topology of four hidden states with no transition

3. EXPERIMENTAL FRAMEWORK
3.1. Database and protocol
We have used for the experiments a subcorpus of the MCYT
bimodal database [18], which includes fingerprint and on-line
signature data of 330 contributors. In the case of the signature
data, skilled forgeries are also available. Imitators are provided the signature images of the client to be forged and, after
an initial training period, they are asked to imitate the shape
with natural dynamics. Signature data were acquired using an
inking pen and paper templates over a pen tablet (each signa-

SE1

SE1

>_XSiR-,s~ I Jr
1 $:l/t1 v=l|StPEr~ ~ ~ ~. . . .
EA

5~5

SE- i2 SE12

9- 3

' E

SEl

'

1l '

,E- 1

SE-33

SE-9

M10

'8E-14

SEz-iSEl.9

8S_..

SE-34

SE-35

T l

SEE

1I]I~~~
- 1[
___ j
7
su2u2

..J..

FE

FE -9

_
_......

Ex

I IXX X3A IEwW

e_

ennn/
........5
su.c............c.s

17

BE

18

MAE

9SE 20 SE 21

SE-22

1111

SE 23

SE

e4

. . ...........
. . . . . . . . ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~a

i dil

_.@@

..FE-3...2

1
_
(L7-

~~~~~~~~~~~~~~...............

S LANT DIRECTION EXTRACTION

.......................................................

ENVELOPE DIRECTION EXTRACTION

Fig. 4. Feature extraction stage performed in the global expert. Structuring elements used for slant direction extraction (SE- I to
SE-32) and envelope direction extraction (SE-33 to SE-38) are also shown. Origin of the element is indicated in gray. The area
of SE-I to SE-32 is 10 pixels and the angle between successive elements is approximately 11 degrees. The areas of SE-33/34
and SE-35/36/37/38 are 7 and 4 pixels respectively.

genuine signatures and 15 forgeries per user (contributed by

3 different user-specific forgers). Examples can be seen in
Figs. 1 and 2.
The experimental protocol is as follows. The training set
comprises either 5 or 10 genuine signatures (depending on
the experiment under consideration). The remaining genuine
signatures are used for testing. For a specific target user,
casual impostor test scores are computed by using the genuine samples available from all the remaining targets. Real
impostor test scores are computed by using the skilled forgeries of each target. As a result, we have 75 x 10 = 750 or
75 x 5 = 375 client similarity scores, 75 x 15 = 1, 125 impostor scores from skilled forgeries, and 75 x 74 x 10 = 55, 500
or 75 x 74 x 5 = 27, 750 impostor scores from random forgeries.
In order to have an indication of the level of performance
with an ideal score alignment between users, results here are
based on using a posteriori user-dependent score normalization [6]. The score normalization function is as follows s' =
s - sx(client, impostor), where s is the raw score computed by the signature matcher, s' is the normalized matching
score and sA ( client , impostor) is the user-dependent decision threshold at a selected point obtained from the genuine
and impostor histograms of user A. In the work reported here,
we record verification results at three points: EER, FAR=1O%
and FRR=1O%.

1 -4244-1 549-7/07/$25.00
2007 IEEE

3.2. Results

All signers in the database used for our experiments are manually assigned a legibility label and a type label. One of three
different legibility labels is assigned: i) name not legible or no
name; ii) uncertain; and iii) name clearly legible. Examples
are shown in Fig. 1. Condition ii) is used in the case that some
characters of the name can be recognized but it is not possible to extract the name completely. In addition, four different
type labels are assigned based on the following criterion: a)
simple flourish; b) complex flourish; c) name + simple flourish; and d) name + complex flourish. Examples are shown in
Fig. 2. It should be noted that signatures of class a) and b) are
those assigned to the non-legible class. Similarly, signatures
of class c) and d) are those assigned to the medium and legible classes. The distributions of signers in the database based
on name legibility and signature type are shown in Table 1.
Table 2 shows the system performance based on name legibility for the two machine experts. Regarding skilled forgeries, we find that the best results are always obtained for the
legible case. The non legible case results in no significant
improvement in most cases or even worse performance with
both machine experts. It could be expected that legible signatures result in worse performance, since they are easier to
imitate, because imitators have some background knowledge
of what they have to imitate. However, it is observed that legible signatures provide better performance than non legible
ones. This may be due to the simplicity of most non-legible

2007 Biometrics Symposium

TR sign
5

10

TRsign
5
10

point
EER
FA=10
FR=10
EER
FA=10
FR=10

point
EER T
FA=10
FR=10
| EER T
FA=10
FR=10

___
Non legible
24.91
FR=45.56
FA=39.81
21.11
FR=38.89
FA=41.29

EXPERT BASED ON GLOBAL INFORMATION

_
_
_
Random forgeries
Skilled forgeries
Medium ] Legible ft Overall
Non legible
Medium ] Legible ] Overall]
26.49
21.58
23.78
8.41
10.58
9.94
9.79
41.47
FR=11.11
FR=13.16
FR=15.53
13.73
FR=44.74
FR=37.63
40.44
FA=13.09
FA=19.06
FA=15 .62
FA=53.68
FA=36.49
15.41
1
|
7.26 1
T
|| 22.13 [
6.57
25.17
20.55
9.47
5.97
38.13
FR=6.67
FR=42.11
FR=36.32
FR=7.89
FR=5.26
6.27
38.4
FA=11.46
FA=13.11
FA=8.50
FA=47.72
FA=32.28
10.32

EXPERT BASED ON LOCAL INFORMATION (HMM)

1
Random forgeries
Skilled forgeries
Overall]
Nonlegible lMedium ] Legible
Medium ] Legible ft Overall
Non legible
| A|4.45
5.21
A45.26 |A5.59
216.54 || 17.76
A|216.67
lA21.23
FR=35.00
FR=39.47
FR=27.37
32.4
FR=1.67
FR=4.21
FR=6.58
4.8
FA=5.62
5.03
FA=24.82
FA=37.19
FA=22.11
26.84
FA=4.14
FA=4.58
|| 2.74
T
|
|| 14.44 |
|
2.28
3.27
20.00
10.61
1.51
16.67
FR=23.33
FR=31.58
FR=18.42
22.93
FR=0.00
FR=1.05
FR=4.74
2.67
FA=4.69
FA=32.63
122.04
FA=1.81
FA=4.35
3.82.7
FA=22.22
FA=16.84

Table 2. System performance based on name legibility. Results are given in %.

signatures.
Regarding random forgeries, we observe from Table 2
that for the expert based on global information, improvement
achieved depends on the number of signatures used for enrolment. When using 5 signatures, the best results are obtained
for the non legible case, whereas when using 10 signatures,
the best results are for the legible signature case. On the other
hand, for the machine expert based on local information, the
best performance is always obtained for the non legible case.
System performance in relation to signature type is shown
in Table 3. Regarding skilled forgeries, Table 2 shows that
non legible signatures resulted in no significant improvement
with either expert. If we divide non legible signatures into
"simple flourish" and "complex flourish", we observe that
complex flourish signatures result in improved performance.
This could be because simple flourish signatures are easier
to imitate than complex flourish ones. It is also worth noting that signatures classified as "name + simple flourish" result in better performance with the global expert, but a worse
performance is obtained with the local expert. The opposite
happens with the "name + complex flourish" samples. This
could be because, since the local machine expert processes
signature images by blocks, it better deals with most complex signatures such as the "name + complex flourish" ones.
In complex signatures, there are regions of the signature image having various strokes crossing in several directions. The
global machine expert is not able to deal satisfactorily with
this case, since it processes the signature image as a whole.
Regarding random forgeries, we observe from Table 3 that
signatures classified as "name + complex flourish" always result in worse performance with both machine experts. Signatures classified as "name simple flourish" result in improved
performance with the global expert, but worse performance is
obtained with the local expert in most cases. The opposite
happens with the "complex flourish" signatures. Also interestingly, simple flourish signatures always work well with the

1-4244-1 549-7/07/$25.OO

2007 IEEE

local expert, but this is not the case with the global expert,
in which the performance becomes poorer as we increase the
number of signatures for enrolment.

4. CONCLUSIONS
In this paper, we evaluate the impact of signature legibility
and signature type on the recognition rates of off-line signature verification systems. For our experiments, we have used
two machine experts that exploit information at two different
levels. The first is based on global image analysis and a statistical distance measure, whereas the second is based on local
image analysis and left-to-right Hidden Markov Models.
Regarding name legibility criteria, similar behaviour is
found for both machine experts for the skilled forgeries experiments. The best results are always obtained for the legible case, whereas the non legible case results in no significant

improvement, or even worse performance.

It could be expected that legible signatures result in worse

performance for skilled forgeries, since they are easier to imitate, however this is not the case in our experiments. Characteristics such as signature complexity or stability could have
clearer impact in the performance [7, 19] and this will be the
target of future work. In our experiments, we observe that
the most complex signatures ("name + complex flourish") are
quite robust to skilled forgeries using the HMM system, although they are not suitable to discriminate between different
signers (i.e. random forgeries). The opposite happens with
the most simple signatures ("simple flourish").
Exploiting differences in performance of several matchers
with respect to a measurable criteria can be used to improve
verification rates, as shown in other biometric traits (e.g. see
[20]). For instance, the steps of the recognition system can
be adjusted or different matchers can be invoked based on the
measured criteria.

2007 Biometrics Symposium

|TR
sign
5

10

|TR
sign
5

10

point

EER
FA=10
FR=10
EER
FA=10
FR=10

point

EER
FA=10
FR=10
EER
FA=10
FR=10

Simple
flourish
26.33
FR=68
FA=37.33
20
FR=48
FA=57.33

|Simple
flourish
- T 25.67
FR=52.00
FA=42.67
- T 25.33
FR=36.00
FA=29.33

Complex
flourish
23.72
FR=36.92
FA=40.77
21.12
FR=35.38
FA=34.87

EXPERT BASED ON GLOBAL INFORMATION

Skilled forgeries
Random forgeries
Simple
Name +
Name +
Name +
Overall
Name +
Overall
Complex
flourish
flourish
simple fl. complex fl.
simple fl. complex fl. _|
]
20.33
28.18
23.78
4.14
10.06
7.24
14.74
9.79
FR=15.38
FR=9.71
FR=22.73
FR=13.73
FR=35.14
FR=47.73
FR=41.47
FR=0.00
FA=15.41
FA=2.89
FA=17.06
FA=8.05
FA=29.21
FA=36
FA=49.70
FA=40.44
22.32
22.41
22.13
6.94
5.70
1
7.261
7.97
9.53
FR=4.00
FR=7.69 1 FR=4.57
FR=8.64
FR=6.27 1
FR=36.57
FR=40.91
FR=38.13
FA=10.32 ]
FA=19.43
FA=8.41
FA=8.68
FA=12.24
FA=35.05
FA=42.12
FA=38.4

Complex
flourish
13.85
FR=28.46
FA=18.72
12.82
FR=18.46
FA=20.00

EXPERT BASED ON LOCAL INFORMATION (HMM)

Random forgeries
Skilled forgeries
Name +
Name +
Name +
Name + 11 Overall
Overall
Simple
Complex
flourish
flourish
simple fl. complex fl. 11
simple fl. complex fl.
]
1
1
1
11
5.21
1
4.00
4.67
4.86
6.41
21.57
12.58
17.76 1
I FR=36.29
1 F FR=2.00 1 FR=1.54 1 FR=5.14
FR=6.82
FR=24.10
32.4
4.8
FA=3.84
FA=4.36
FA=4.90
FA=6.10
FA=33.52
FA=17.58
26.84 ]
5.03
1 2.741
0.03
2.08
1.71
4.84
15.33
11.82 1
14.441
I FR=25.71
1
2.67
FR=0.00 1 FR=3.43
FR=3.64 11
FR=18.18 11 22.93 1 F FR=0.00
FA=0.22
FA=2.39
FA=2.72
FA=7.26
3.82
FA=22.48
FA=21.21
22.04 ]

Table 3. System performance based on signature type. Results are given in %.

5. REFERENCES
[1] M.C. Fairhurst, "Signature verification revisited: promoting
practical exploitation of biometric technology," Electronics

prandcommunexplicationEngbinmeering Journal,vol.9," pp.e273

[2]

[3]
[4]
[5]

[6]

[7]

[8]
[9]

[10]
[11]

December 1December
a997.
1997.
A.K. Jain, A. Ross, S. Prabhakar, "An introduction to biometric recognition," IEEE Trans. Circuits and Systems for Video
Tech., vol. 14, no. 1, pp. 4-20, 2004.
G. Dimauro et al., "Recent advancements in automatic signature verification," Proc. IWFHR, pp. 179-184, 2004.
R. Plamondon and S.N. Srihari, "On-line and off-line handwriting recognition: A comprehensive survey," IEEE Trans.
on PAMI, vol. 22, no. 1, pp. 63-84, 2000.
D.Y. Yeung et al., "SVC2004: First international signature
verification competition," Proc. ICBA, Springer LNCS-3072,
pp. 15-17, July 2004.
J. Fierrez-Aguilar, J. Ortega-Garcia, and J. GonzalezRodriguez, "Target dependent score normalization techniques
and their application to signature verification," IEEE Trans.
SMC-C, vol. 35, no. 3, 2005.
C. Allgrove and M.C. Fairhurst, "Enrolment model stability in
static signature verification," in in Proc. IWFHR, pp. 565-570,
2000.
J.J. Brault, R. Plamondon, "A complexity measure of handwritten curves: Modeling of dynamic signature forgery," IEEE
Trans. SMC, vol. 23, pp. 400-413, 1993.
M.C. Fairhurst and E. Kaplani, "Perceptual analysis of handwritten signatures for biometric authentication," IEE Proc.
VISP, vol. 150, pp. 389-394, 2003.
L.L. Lee and M.G. Lizarraga, "An off-line method for human
signature verification," in Proc. ICPR, 1996, p. 195198.
J. Fierrez-Aguilar, N. Alonso-Hermira, G. Moreno-Marquez,
and J. Ortega-Garcia, "An off-line signature verification system based on fusion of local and global information," in Proc.
BIQAW, Springer LNCS-3087, 2004, pp. 295-306.

1-4244-1 549-7/07/$25.OO

2007 IEEE

[12] E. Justino, F. Bortolozzi, R. Sabourin, "Off-line signature verification using HMM for random, simple and skilled forgeries,"
Proc. ICDAR, pp. 1031-1034, 2001.

[13] N. Otsu, "A threshold selection method for gray-level his-

tograms," IEEE Trans. on SMC, vol. 9, pp. 62-66, December

~~~~~~~~~~1979.

an 2.E
14R..onzalez
Addison-Wesley, 2002.
[15] S. Theodoridis and K. Koutroumbas, Pattern Recognition,
Academic Press, 2003.
[16] L.R. Rabiner, "A tutorial on hidden markov models and selected applications in speech recognition," Proceedings of the
IEEE, vol. 77, pp. 257-286, 1989.
[17] J. Ortega-Garcia, J. Fierrez-Aguilar, J. Martin-Rello, and
J. Gonzalez-Rodriguez, "Complete signal modelling and score
normalization for function-based dynamic signature verification," Proc. AVBPA, Springer LNCS-2688, pp. 658-667, 2003.
[18] J. Ortega-Garcia et al., "MCYT baseline corpus: a bimodal
biometric database," IEE Proc. VISP, vol. 150, no. 6, pp. 395401, December 2003.
[19] M.C. Fairhurst, E. Kaplani, and R.M. Guest, "Complexity
measures in handwritten signature verification," Proc. UAHCI,
pp. 305-309, 2001.
[20] J. Fierrez-Aguilar and Y. Chen and J. Ortega-Garcia and A.K.
Jain, "Incorporating image quality in multi-algorithm fingerprint verification," Proc. ICB, Springer LNCS-3832, pp. 213220, 2006.

2007 Biometrics Symposium